jyapayne/worm-scraper
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Fix many erroneously-italicized closing quotes

Browse source
This commit is contained in:
Domenic Denicola 2020-11-14 18:13:12 -05:00
commit d2402b1bb2
2 changed files with 31 additions and 17 deletions
Download patch file Download diff file Expand all files Collapse all files

8
lib/convert-worker.js
View file

@ -161,9 +161,9 @@ function getBodyXML(chapter, book, contentEl) {
xml = xml.replace(/<em>‘\s*([^<]+)\s*’<\/em>/g, "‘<em>$1</em>’");
xml = xml.replace(/<em>‘\s*([^<]+)\s*<\/em>\s*’/g, "‘<em>$1</em>’");
xml = xml.replace(/‘\s*<em>\s*([^<]+)\s*’<\/em>/g, "‘<em>$1</em>’");
xml = xml.replace(/<em>“\s*([^<]+)\s*”<\/em>/g, "“<em>$1</em>”");
xml = xml.replace(/<em>“\s*([^<]+)\s*<\/em>\s*”/g, "“<em>$1</em>”");
xml = xml.replace(/“\s*<em>\s*([^<]+)\s*”<\/em>/g, "“<em>$1</em>”");
xml = xml.replace(/<em>“\s*([^<”]+)\s*”<\/em>/g, "“<em>$1</em>”");
xml = xml.replace(/<em>“\s*([^<”]+)\s*<\/em>\s*”/g, "“<em>$1</em>”");
xml = xml.replace(/“\s*<em>\s*([^<”]+)\s*”<\/em>/g, "“<em>$1</em>”");
xml = xml.replace(/([^\n>])<em> ?/g, "$1 <em>");
xml = xml.replace(/ ?<\/em>/g, "</em> ");
xml = xml.replace(/<p([^>]+)> <em>/g, "<p$1><em>");
@ -197,6 +197,8 @@ function getBodyXML(chapter, book, contentEl) {
xml = xml.replace(/‘<em>([^<]+)<\/em>‘/g, "‘<em>$1</em>’");
xml = xml.replace(/I”m/g, "I’m");
xml = xml.replace(/<em>([a-z]+)!<\/em>/g, "<em>$1</em>!");
xml = xml.replace(/(?<! {2})<em>([\w ’]+)([!.?])”<\/em>/g, "<em>$1</em>$2”");
xml = xml.replace(/<em>([\w ’]+[!.?])”<\/em>/g, "<em>$1</em>”");
// Similar problems occur in Ward with <b> and <strong> as do in Worm with <em>s
xml = xml.replace(/<b \/>/g, "");